User:ONUnicorn/thoughts on deletion

Random thoughts on deletion and deletion processes.

Speedy Deletion
It amazes me how few of the speedies I decline are subsequently nominated for AFD (or MFD if not in article space). A decline of a CSD nomination should not be the final word on if the article belongs in Wikipedia or not. The Criteria for Speedy Deletion are supposed to be several categories of pages where the community has decided that these pages are so toxic, or so obviously inappropriate, that they can be summarily executed with no further discussion. The criteria are to be read strictly, and even if one does not think the page belongs, if it doesn't meet the criteria, it should be declined and a different process should be used.

Sometimes it seems like new page patrollers are extremely reluctant to use the AFD process, so they try to shoehorn everything into CSD criteria or draft space. But CSD is supposed to be only for the most obvious cases meeting narrowly tailored criteria.

G1 - Patent nonsense
Before I was an admin I had a couple G1s declined, which helped calibrate my bar for G1. In order to qualify for G1, the page has to literally not make any sense. For example:
 * A page with the text: "orfwkeiorhngrtgrojinf ewi ofj e ruwhfhewrrfewhjo9ewijfweonf" qualifies.
 * A page with text like: "Faster they ran. Outer ground up the hill. Shakespeare play faster make now." qualifies.
 * A page with text like: "Charlie was a dragon. He lived in the 1700s some time, idk exactly. He knew Marie Curie.  And Mari Kondo. And everyone. He was awesome." does not qualify because it makes literal, grammatical sense. It is coherent and understandable.  This does not mean that a page like that couldn't be deleted under some other criteria; such as A11 if it's in article space, or U5 if it's in userspace and meets the other criteria, possibly G2 or G3, depending on context.  But it's not G1.

G2 - Test pages
To delete a page as a test page, I want there to be some clear indication that it was created to test Wikipedia's editing functions. For example, My history page was created with the edit summary, "Content testing", therefore G2 clearly and unambiguously applies. Likewise Tesztrepter had text indicating that it was created as a test page.

I have deleted some as G2 where it was less clear. In those instances I'm also looking at the contributor's other edits for some indication that they are experimenting rather than making a serious attempt at a draft or an article or whatever namespace it's in. Many things get tagged as G2 that are probably more properly tagged as G3 or A11; but the tagger is trying to assume good faith by calling it a test rather than vandalism. This is similar to the way uw-test1 is often used rather than uw-vandalism1. If I think this is the case, I'll not change the criteria before deleting, and go ahead with the gentler, AGF reason.

G3 - Pure vandalism and blatant hoaxes
Why are these lumped together? It seems vandalism and hoaxes should have separate criteria. Then again, hoaxes are a subset of vandalism.

RE: Pure vandalism

 * A lot of this could conceivably fall under G2 (test pages) and G10 (attack pages).

RE: Hoaxes

 * I think it's important for hoaxes that the criteria is limited to blatant hoaxes. To some extent, I think a blatant hoax is an oxymoron. Blatant means "obvious", but a hoax is defined as "an attempt to trick an audience into believing that something false is real."  If it's really that obvious, it won't trick anyone.
 * The more research I have to do to determine if it is a hoax or not, the less inclined I am to delete it under this criteria, as if I'm having to do substantial research, it's not blatant.
 * Sometimes I'll ask for more information from the person who tagged it. That usually reassures me that it can/should be deleted under this criteria.

G4 - Recreation of previously deleted content
See User:ONUnicorn/G4. I was using that page to gather thoughts and evidence a few years ago for the RHaworth arbcom case. Although the evidence specific to that case is rather moot at this point, some of the thoughts and points recorded there remain valid.

G5 - Creations by banned or blocked users
I have mixed feelings about this criteria. Obviously, people get blocked or banned for good reasons, and we want to discourage sockpuppetry. Hence the reasoning behind this criteria. However, if the content is otherwise acceptable (i.e. the topic is notable, the article cites multiple reliable, independent sources, the information is presented from a neutral point of view, the article is not in need of a total rewrite, etc.), it seems a shame to delete it.

When I'm evaluating an article for CSD; I want to be evaluating content, not judging contributors. Moreover, I am not a checkuser (and that's one Wikipedia position I have no interest in), so I have no way of verifying that the article was created via sockpuppetry. I can look to see if the creator has been blocked as a sockpuppet, but I cannot independently verify that it was the same person.

I think the saving grace of this criteria is the "and that have no substantial edits by others" language. If someone other than the creator has taken the page in hand and expanded it, then they are at least partially responsible for the content and it can stay.

G7 - Author requests deletion
This is pretty simple. Check the history. Did the author request deletion? Is there a substantial contribution by someone else? If yes to the first and no to the second, delete. Sometimes it can get a little ambiguous as to if the person requesting deletion is the author of the majority of the content, but most of the time it's straight forward.

The thing about this though, is often it is an indication of someone who is deeply frustrated with Wikipedia and our processes. Take Draft:Lauren Burch for example. The draft creator included the following comment when tagging it for deletion, "I wanted to create a page of her, but I want it deleted, because I feel like Wikipedia is way too complicated for me". Between creation and then it had been declined at AFC for insufficient sources twice. They had asked the first reviewer how to determine what websites are reliable sources and what websites are not, but had not gotten a response.

G11 - Unambiguous advertising or promotion
Didn't this once say "Blatant advertising"? I think "Unambiguous" is a lower bar than "Blatant", but either way I think it needs to read like an ad (or a PR piece, which can be more subtle). I think a lot of people tagging are assuming an intent on the part of the article/draft writer (in violation of AGF), rather than evaluating the text that is before them on the screen. It seems like companies and BLPs are almost tagged G11 reflexively, no matter how neutrally written they are. I once saw an article about a defunct company that went out of business in the late 1800s tagged as G11. One of the main things I'm looking for is a call to action, "Buy this product!" "Vote for this candidate!" "Support this cause!" "Visit this website!" Another thing I'm looking for is something that looks like an attempt to whitewash a bad reputation, or to create the appearance of a good reputation where no reputation actually exists.

Also; in my mind the bar for what constitutes advertising to the extent that G11 is appropriate is set slightly higher for draft space than main space. In other words, I'm more tolerant of borderline cases in draft than in article space. That doesn't mean I won't delete blatant ads in draft space; just the bar for how blatant is a little different.

Likewise, there is a higher bar in my mind for G11 in cases where a company is no longer in business or a person is deceased. No one is out there trying to get people to visit a restaurant that is closed or campaign for a dead person to be elected to office. Now, there may be some promotion for deceased authors, musicians, and actors where there is some estate or studio or label or publisher that has retained rights to their work and is still trying to sell or promote it, and there may be some promotion by second-hand sellers of products of defunct manufacturers, so it's not impossible for articles on deceased individuals or defunct companies to be promotional, but it's far less likely so my bar for deleting those is higher.

Another issue I see with G11 is that often an article will get tagged as G11 because someone sees some instances of peacock language. Peacock language definitely can be an indication of promotion, but I don't think it always is something that cannot be fixed. G11 only applies if the page is " promotional and would need to be rewritten to serve as (an) encyclopedia article( s ), rather than (an) advertisement( s )". If the page in question can be edited to change, "X is a famous actor at the top of his craft" to "X is an actor", it is not G11 material.

G13 - Abandoned drafts
I don't do much G13 deletion. I get the point of G13 and wanting to clean out abandoned drafts, but from the standpoint of someone who always has 5 zillion half-finished projects lying around with the full intention of getting back to them eventually, I find G13 annoying, and I'm sure newer editors who think they can work on drafts in peace find it annoying too. And yes, you are supposed to be able to get your G13'd draft back at WP:REFUND, but that's also annoying - when you request undeletion you have no way of knowing how long that's going to take, and you have a few hours to work on your draft available now but maybe not tomorrow when it's been undeleted.

It seems like we need a better way of managing stale drafts. Everything about draftspace is well-intentioned, and horribly broken. G13 is no exception.

A7 - Assertion of importance
A7 does not require a showing of notability, merely an "indication of importance". An indication of importance is a lower bar than notability. An indication of importance is "Neil Armstrong was the first person to walk on the moon". Notability is, "Neil Armstrong was the first person to walk on the moon.  " An article can be completely unsourced, unsourcable, and unnotable, yet still pass A7. The above is taken from User_talk:ONUnicorn/archived_talk_9. A lot of people tagging for A7 seem to confuse "indication of importance" with notability; they are not the same and "indication of importance" is intentionally a MUCH lower bar than notability. Continuing with the Neil Armstrong example above, a version that would be deletable under A7 would be a version that just said (in its entirety), "Neil Armstrong was born in Ohio."

A10 - Recently created article that duplicates an existing topic
I'm really not 100% sure why this is even a speedy criteria. IMO, 90% of the time, it's better to change the recently created article into a redirect than delete it. Duplicate articles get created for a reason, and that reason usually falls into one of two categories: Either way, a redirect fixes any issues of people searching at the wrong title, while still preserving any information or sources in the redirect history in case there is something that can or should be merged.
 * 1) The person could not find the article due to their use of a different title, and created the article because they thought there wasn't one.
 * 2) It's a content fork. The person didn't like what the other article had to say and couldn't get consensus for their version.

The remaining 10% that should be deleted are actually covered by other existing criteria. Content forks are often created when the person does not like the content in the original article - either they do not like the subject and think the article is unbalanced in the subjects favor, so they create what is essentially an attack page, or they think it is unbalanced with too much criticism so they create a promotional version. Either way, those should be deleted, but they have their own criteria. Sometimes the new creation is a copyvio - which again has its own criteria.

Most of the time if I come accross something tagged as A10, I'll redirect it rather than deleting unless it meets some other speedy criteria.

Prod
A lot of people think Prod is broken. I kind of think it's nice. If an article doesn't fit a CSD category, but you don't want to go through the hassle of an AFD, prod it. If the article creator isn't active, it'll go away. If someone doesn't think it should be deleted, they can remove the tag and either fix it or go through an AFD. Prods can also be restored at REFUND, like G13.

Cleanup
We often hear that "AFD is not cleanup", but for all practical purposes, it is. Wikipedia's mechanisms for fixing sub-par but notable content are almost dysfunctional. We have a great system of flagging issues via clean up tags, but getting people to actually take action to fix the tagged issues is - difficult. There are (as of the time of writing this) approximately 134,000 articles tagged as not having a single reference, and nearly half a million with at least one citation needed tag. There are thousands needing a copy edit at any given time. There are 118,000 articles with multiple issues.

Sticking articles in massive maintenance categories allows the person tagging the articles to feel like they've done something - they've flagged the issue. But it doesn't prompt people to actually fix the tagged issue. There are coordinated campaigns to focus on various maintenance categories, but participation tends to be spotty and they are overwhelmed. All this leads some people to think that it's better to delete poor articles rather than leave them be in the hope that someone will eventually fix them. However, that conflicts with the wiki model.

Nominating these articles for deletion will often prompt people to fix them. See The Heymann Standard for an example. However, a lot of people are discouraged by the fact that the article has been nominated for deletion. I've seen comments in deletion discussions to the effect of, "This is fixable, but I don't want to waste my time fixing it if it's just going to get deleted." They think they need to wait for the discussion to end to take action.

Then, once the discussion is closed and there is no longer an axe hanging over the article's head, they lose interest in fixing it.

I think we need some kind of new process. A place where problematic articles can be discussed and people can try to clean them up. There needs to be a time limit - but I'd like it to be closer to a month than the week that is the time limit at AFD. There needs to be some incentive, some kind of axe hanging over the article to get people to actually work on it. So I'd like the default to be that if the article is not improved after a month (or whatever gets decided on), then off to AFD with it. I'd like this to be an optional pre-AFD step, and if someone skipped this step and nominated straight to AFD, I'd like "send for cleanup" to be a valid result of an AFD. I'd also like "send to AFD" to be a valid result of this process. It needs to be different enough from existing clean up processes that it'll serve as a compliment to them (rather than competition), and accomplish what they haven't.

Relisting

 * Articles for deletion/Log/2023 December 4 62 AFDs
 * Articles for deletion/Log/2023 December 5 124
 * Articles for deletion/Log/2023 December 6 79
 * Articles for deletion/Log/2023 December 7 108
 * Articles for deletion/Log/2023 December 8 98
 * Articles for deletion/Log/2023 December 9 100